Students and professionals active in the area of environmental sciences and water resources.
Geostatistics nowadays is met to many commercial software, while toolboxes have been developed in various programming environments. However, the users often ignore the basics of proper geostatistical analysis. This course aims to introduce the basic geostatistical theory using applications in environmental science and to introduce the recent advancements for improved spatiotemporal data analysis.
Geostatistics is closely related to water resources and environmental science due to the management of spatial observations to produce the spatial variability of measurement variables. It is usually applied in mapping of spatial observations like from monitoring stations for hydrological or water resources data, assessing spatial data quality, relating the accuracy of spatial data to their intended use, sampling design optimization, modelling dependency structures, and drawing valid inference from a limited set of spatial data in agriculture, hydrology, hydrogeology, soil science, ecology e.t.c.
The first part of this course starts with statistical and geostatistical inference of data with known locations presenting several interpolation methods. Then, the theory of spatial variability is explained presenting the structure of spatial variability-dependence with covariance functions and variograms. Data normalization techniques and trend analysis is presented for appropriate geostatistical applications. Comparison of geostatistical methodologies is applied, optimal sampling is discussed and attention is further given to multivariate geostatistics.
The second part of the course considers modelling spatiotemporal phenomena which is an important topic in today’s research. However, the extension from pure spatial to spatiotemporal approaches is not trivial. Spatiotemporal covariance models (separable, non-separable) will be presented and explored on real datasets. The practical part deals with the exploration of different spatiotemporal variogram models and space-time kriging approaches applied to an own data set or a provided example.
Upon completion, the participant should be able to:
– Perform proper and efficient sample statistical assessment and to statistically characterize spatially referenced data
– Know the advantages and disadvantages of stochastic and deterministic geostatistical techniques and to appropriately select and apply the right geostatistical approaches
– Apply effective quantitative analysis of spatial and spatio-temporal data
– Work with real hydrological data and produce efficient and useful maps
– Work comfortably in R programming environment for statistics
– Should be able to choose and describe adequate spatial and spatiotemporal continuity models (variograms) for different applications